0

Here is the problem:

enter image description here

When cheching the Chi squared distribution table, the it seems like in the solution the denominators should be switched, because for .025 quantile the value is 13.844 and for .975 it is 41.923. But here everything is the other way round. Can someone explain?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
  • The solution clearly places the endpoints of the interval in the correct order. (If algebra doesn't make that obvious, then do the calculations to see.) What, then, is your question?? – whuber Nov 27 '20 at 23:05
  • In many statistics books the subscripts used on notation to cut 5% of the probability from the upper tail of a distribution are $t_{.05}, \chi^2_{.05},$ etc. These are sometimes called (upper) "percentage points" of the relevant distribution. In terms of percentiles they are, of course, 95th percentiles. // So, for degrees of freedom $\nu=10$ you cut 5% from the upper tail of the chi-squared distribution with R code `qchisq(.95,10)`, which returns $18.30704.$ But in a printed table you'll look in a column headed $\chi^2_{0.5}$ to find $18.31.$ – BruceET Nov 28 '20 at 00:04
  • @whuber But I don't understand, if confidence level is 0.05, then $\alpha / 2$ = 0.025 and in Chi square distribution table the corresponding value for 0.025 is 13.844. – Gianni D'Adova Nov 28 '20 at 08:19
  • [This search](https://stats.stackexchange.com/search?q=confidence+two+tail*) links to answers to that implicit question – whuber Nov 28 '20 at 15:12

0 Answers0